The nucleoporin-like protein NLP1 (hCG1) promotesCRM1-dependent nuclear protein export

Inga Waldmann, Christiane Spillner and Ralph H. Kehlenbach*Department of Biochemistry I, Faculty of Medicine, Georg-August-University of Gottingen, Humboldtallee 23, 37073, Gottingen, Germany

*Author for correspondence (rkehlen@gwdg.de)

Accepted 25 July 2011Journal of Cell Science 125, 144–154� 2012. Published by The Company of Biologists Ltddoi: 10.1242/jcs.090316

SummaryTranslocation of transport complexes across the nuclear envelope is mediated by nucleoporins, proteins of the nuclear pore complex thatcontain phenylalanine-glycine (FG) repeats as a characteristic binding motif for transport receptors. CRM1 (exportin 1), the major

export receptor, forms trimeric complexes with RanGTP and proteins containing nuclear export sequences (NESs). We analyzed the roleof the nucleoporin-like protein 1, NLP1 (also known as hCG1 and NUPL2) in CRM1-dependent nuclear transport. NLP1, whichcontains many FG repeats, localizes to the nuclear envelope and could also be mobile within the nucleus. It promotes the formation of

complexes containing CRM1 and RanGTP, with or without NES-containing cargo proteins, that can be dissociated by RanBP1 and/orthe cytoplasmic nucleoporin Nup214. The FG repeats of NLP1 do not play a major role in CRM1 binding. Overexpression of NLP1promotes CRM1-dependent export of certain cargos, whereas its depletion by small interfering RNAs leads to reduced export rates.

Thus, NLP1 functions as an accessory factor in CRM1-dependent nuclear protein export.

Key words: NLP1, hCG1, Nuclear export, CRM1, Nucleus, Nup214

IntroductionExchange of molecules between the cytoplasm and the nucleus

occurs through nuclear pores that are embedded in the nuclearenvelope. The nuclear pore complex (NPC) is a giant protein

assembly with many copies of approximately 30 individualnucleoporins (Nups) [for a review see Wente and Rout and

references therein (Wente and Rout, 2010)]. Four transmembraneNups that anchor the NPC in the double lipid bilayer of the

nuclear envelope have been described in vertebrate cells (Chadrinet al., 2010; Gerace et al., 1982; Hallberg et al., 1993; Mansfeld

et al., 2006; Stavru et al., 2006). Approximately 15 differentstructural Nups form subcomplexes that participate in the

formation of characteristic NPC building blocks such as thenuclear basket and nuclear or cytoplasmic rings. Finally,

approximately 10 Nups are rich in phenylalanine–glycine (FG)

repeats, motifs that mediate the interaction of the NPC withsoluble nuclear transport receptors. This interaction is the basis of

all models of the mechanisms for the selective translocation ofmacromolecules across the NPC (for reviews, see Goldfarb,

2009; Terry and Wente, 2009; Walde and Kehlenbach, 2010;Wente and Rout, 2010). The majority of the soluble factors

belong to the superfamily of importin-b-like transport receptors.They are called importins or exportins, depending on the major

direction of cargo transport, and are collectively also referred toas karyopherins (for a review, see Fried and Kutay, 2003). The

prototype importin is importin-b itself, which, through theadaptor protein importin-a, interacts with proteins containing a

classical nuclear localization signal (cNLS). The major exportin

is exportin 1, better known as CRM1 (Fornerod et al., 1997a;Fukuda et al., 1997; Kehlenbach et al., 1998; Neville et al., 1997;

Ossareh-Nazari et al., 1997; Stade et al., 1997). CRM1 interactswith proteins containing a nuclear export sequence (NES) and

mediates transport of hundreds of different proteins and certainRNAs out of the nucleus (for a review, see Hutten and

Kehlenbach, 2007). The first CRM1-dependent NESs wereidentified in the HIV-1 Rev protein, which serves as an adaptor

for nuclear export of unspliced viral RNAs (Fischer et al., 1995)and in the inhibitor of the catalytic subunit of cAMP-dependent

protein kinase, PKI (Wen et al., 1995). Recently, the crystalstructure of CRM1 in a complex with export cargos was

determined (Dong et al., 2009; Monecke et al., 2009). An

important component of export complexes is the small GTP-binding protein Ran in its GTP-bound form, which binds

cooperatively to CRM1, together with export cargos (Fornerodet al., 1997a). Similar to nuclear import complexes, CRM1-

containing export complexes can interact with FG Nups. Amongthese is Nup214, a nucleoporin that localizes to the cytoplasmic

side of the NPC and that is required for efficient export of some,but not all, CRM1 cargos (Bernad et al., 2006; Bernad et al.,

2004; Hutten and Kehlenbach, 2006). More recently, Nup98was described as a CRM1-interacting nucleoporin that promotes

nuclear export (Oka et al., 2010).

In this study, we investigated the role of the ill-defined

nucleoporin-like protein NLP1 (also known as NUPL2) inhumans, which has a presumptive ortholog, Rev-interacting

protein (Rip1p; also known as Nup42), in yeast. Rip1p wasoriginally identified in a yeast two-hybrid screen, searching for

yeast proteins that bind to the viral protein HIV-1 Rev (Stutzet al., 1995). Later, the human protein human candidate gene 1

(hCG1) was described (Van Laer et al., 1997). hCG1 has 55%

amino acid sequence homology to Rip1p over the entire length ofthe proteins (423 aa in humans) and can substitute for Rip1p

functions in yeast (Strahm et al., 1999). hCG1 was also identifiedas NLP1 (nucleoporin-like protein 1), again as a protein

144 Research Article

Journ

alof

Cell

Scie

nce

interacting with HIV-1 Rev in a two-hybrid screen (Farjot et al.,

1999). A characteristic of the yeast and the human proteins is

the abundance of FG motifs, which are also found in FG

nucleoporins. Although the yeast protein Rip1p is not essential, it

seems to be required for export of certain heat shock mRNAs

(Saavedra et al., 1997; Stutz et al., 1997). Another study

suggested a role of Rip1p in nuclear export of heat shock and

non-heat shock mRNAs at elevated temperatures (Vainberg et al.,

2000). Rip1p was found to also interact with CRM1 in two-

hybrid assays (Neville et al., 1997), as well as in vitro using

purified proteins (Floer and Blobel, 1999). A Rip1p-deletion

strain did not exhibit a profound defect in CRM1-dependent

nuclear protein export (Stade et al., 1997). Hence, the functional

relevance of Rip1p–CRM1–RanGTP complexes (Floer and

Blobel, 1999) remains unclear. The human protein NLP1

interacts with the mRNA export factor TAP (Katahira et al.,

1999) and functions in nuclear export of Hsp70 mRNA (Kendirgi

et al., 2005). Similar to the yeast protein, an interaction between

NLP1 and CRM1 was detected in two-hybrid assays (Farjot et al.,

1999). Nothing is known, however, about the function of NLP1

in nuclear export of protein cargos in mammalian cells. Here, we

describe a supportive role of NLP1 in CRM1-dependent export.

ResultsCRM1 concentrations are rate-limiting for nuclear export of

artificial reporter proteins

The cellular concentration of nuclear import factors of the

importin-b superfamily is rate-limiting for efficient transport of

proteins into the nucleus (Yang and Musser, 2006). We

investigated whether the concentration of CRM1, the major

nuclear export factor, is also rate-limiting, and analyzed the

subcellular localization of nuclear shuttling proteins in control

HeLa cells and in cells overexpressing CRM1. As a reporter

protein, we used an artificial shuttling protein, a double GFP

containing an N-terminal NES and a C-terminal cNLS (NES–

GFP2–cNLS). This protein localized predominantly in the

nucleus in ,70% of transfected control cells (Fig. 1A,B).

Strikingly, co-transfection of hemagglutinin-tagged CRMI

(CRM1–HA) resulted in a clear shift of the reporter protein

towards the cytoplasm, suggesting that nuclear export of

overexpressed NES–GFP2–cNLS is limited by the cellular

CRM1 concentration. In the presence of leptomycin B (LMB),a selective CRM1 inhibitor (Wolff et al., 1997), NES–GFP2–

cNLS, is localized entirely in the nucleus in control cells and incells overexpressing CRM1–HA, indicating that nuclear exportof the reporter protein indeed involves CRM1 as an export factorand also that its nuclear import was not inhibited under our

experimental conditions. Very similar observations weremade with another shuttling protein, the negative cofactor 2 b(NC2b; see supplementary material Fig. S1A,B). NC2b employs

importin-a and/or -b and importin 13 as import- and CRM1as export receptors, respectively (Kahle et al., 2009). Notably,under conditions of overexpression of CRM1 substrates,

the cytoplasmic localization of the endogenous CRM1 cargoRanBP1 was not compromised, suggesting that the CRM1 systemwas not completely overloaded (supplementary material Fig.S1C). Together, these results demonstrate that in HeLa

cells, CRM1 concentrations are not saturating for export ofoverexpressed nucleocytoplasmic shuttling proteins. We nextanalyzed the function of NLP1, a putative CRM1-binding

protein, as a potential accessory factor that might stimulatenuclear protein export under certain conditions.

NLP1 interacts with CRM1 in a RanGTP- and exportcargo-dependent manner

The yeast ortholog of NLP1, Rip1p/Nup42p, forms trimericcomplexes with CRM1 and RanGTP. Export cargos such as HIV-

1-Rev, however, appeared to be excluded from such complexes(Floer and Blobel, 1999). In yeast two-hybrid assays, humanNLP1 has previously been described to interact with CRM1

(Farjot et al., 1999). A direct binding or a functional role innuclear export, however, has not been reported. We therefore setout to analyze the binding of NLP1 to CRM1 using recombinant

proteins, and to investigate the role of this protein in CRM1-dependent export. First, we expressed NLP1 fused to the maltosebinding protein (MBP) and immobilized it on beads. CRM1 and

RanQ69L, a Ran mutant that is insensitive to the GTPase-activating protein RanGAP (Klebe et al., 1995), loaded with GTPwere then added in combination with an NES peptide or withGST–snurportin 1 (GST–SPN1) as a CRM1-dependent export

cargo (Dong et al., 2009; Monecke et al., 2009; Paraskeva et al.,1999). Very little binding of CRM1 to NLP1 was observed in theabsence of RanQ69L–GTP, whether or not an export cargo was

present in the reaction (Fig. 2A, lanes 1, 2, 5, 6). Addition ofRanQ69L–GTP alone resulted in detectable levels of CRM1binding to NLP1 (lanes 3, 7). Including either GST–SPN1 (lane

4) or the NES–peptide (lane 8) in the reaction further increasedCRM1 binding to MBP–NLP1, suggesting the formation oftetrameric complexes. Binding of GST–SPN1 was only detectedwhen RanGTP was present in the reaction (compare lanes 2

and 4). Importantly, CRM1 and RanQ69L–GTP bound toimmobilized MBP–NLP1 but not to MBP alone when addedto the reaction together with an export cargo (lanes 9 and

10), demonstrating the specificity of the interaction. In acomplementary experiment, we immobilized GST–Ran–GTPand analyzed binding of CRM1, MBP–NLP1 and SPN1. Again,

the strongest signals were observed when CRM1 was addedtogether with MBP–NLP1 and His–SPN1 (Fig. 2B, lane 6). LessCRM1 was detected when either His–SPN1 (lane 4) or MBP–

NLP1 (lane 5) was omitted from the reaction. Only backgroundbinding was detected when the immobilized GST–Ran had beenloaded with GDP (data not shown) or when unfused GST had

Fig. 1. CRM1 concentrations are rate-limiting for nuclear protein

export. HeLa cells were co-transfected with plasmids coding for CRM1–HA

or an empty control vector and NES–GFP2–cNLS and treated with LMB

(5 nM) for 2.5 hours as indicated. (A) Microscopic analysis. Scale bar:

10 mm. (B) Quantification of cells with a predominant nuclear signal (N.C)

of the reporter protein. Error bars show the standard deviation of the mean of

three (+ LMB) or six (–LMB) independent experiments.

NLP1 in CRM1-dependent export 145

Journ

alof

Cell

Scie

nce

been immobilized (Fig. 2B, lane 8). To analyze the interaction of

CRM1–NLP1 complexes with RanGTP and RanGDP in a more

quantitative manner, we performed an ELISA-type binding assay.

GST–Ran loaded with either GDP or GTP was bound to

microtiter plate wells and incubated with MBP–NLP1 and

increasing concentrations of CRM1. Bound NLP1 was thendetected with a specific anti-NLP1 antibody. As shown

in Fig. 2C, only background levels of NLP1 binding toimmobilized RanGDP were observed, even at the highestCRM1 concentration. With immobilized RanGTP, by contrast,increasing CRM1 concentrations resulted in increased binding of

NLP1. Especially at lower CRM1 concentrations, the addition ofHis–SPN1 to the reaction further promoted binding of NLP1(together with CRM1) to RanGTP. With or without SPN1, half-

maximal binding was observed at CRM1 concentrations in thelow nanomolar range. These results show that NLP1 can formtrimeric complexes with RanGTP and the export receptor CRM1.

CRM1 cargos such as SPN1 or an NES peptide can join thesecomplexes, suggesting that NLP1 functions in CRM1-mediatednuclear export rather than in re-import of CRM1 back into thenucleus.

We previously showed that wild-type RanGTP in a complexwith CRM1 and the nucleoporin Nup214 is resistant to RanGAP-promoted GTP hydrolysis (Hutten and Kehlenbach, 2006).

Likewise, a ternary complex of yeast CRM1, RanGTP andRip1p/Nup42p was protected against RanGAP (Floer andBlobel, 1999). In RanGAP assays, a C-terminal fragment of

Nup214 strongly reduced the ability of RanGAP to promote GTPhydrolysis on Ran in the presence of CRM1 (Fig. 3A),confirming our previous results (Hutten and Kehlenbach, 2006).Similarly, the addition of NLP1 to the reaction inhibited GTP

hydrolysis. By contrast, Nup88, a nucleoporin that does notinteract with CRM1, was not active in this assay, as shown before(Hutten and Kehlenbach, 2006). This result supports the notion

that trimeric complexes containing RanGTP, CRM1 and NLP1can form. In a late step of transport, such trimeric complexes (andof course tetrameric ones containing an export cargo) have to

dissociate. The RanGTP-binding protein RanBP1 has previouslybeen implicated in terminal steps of nuclear export (Kehlenbachet al., 1999). We therefore tested the ability of RanBP1 to

disassemble NLP1-containing complexes. As before, tetramericcomplexes containing immobilized MBP–NLP1, CRM1,RanQ69L–GTP and an NES-peptide were assembled. After thereaction, RanBP1 was added, and binding of CRM1 and Ran to

NLP1 was analyzed. Clearly, RanBP1 dissociated the pre-assembled complexes because the levels of bound CRM1 andRan were strongly reduced (Fig. 3B). Very similar observations

were made with SPN1 as a CRM1 cargo (data not shown).Together, our results demonstrate the ability of NLP1 to interactwith CRM1 and RanGTP, with or without export cargos,

suggesting a role in nuclear protein export. Such complexes areexpected to be stable until they are dissociated by the concertedaction of cytoplasmic RanBP1 and RanGAP.

Nup214 is a well-characterized binding partner of CRM1 and

is involved in nuclear export of a subset of proteins (Bernad et al.,2006; Fornerod et al., 1997a; Hutten and Kehlenbach, 2006).According to our previous results, it functions at a late stage of

nuclear export, i.e. before disassembly of the RanBP1-mediatedexport complex (Hutten and Kehlenbach, 2006; Kehlenbachet al., 1999). We therefore asked whether NLP1 and Nup214

interact with CRM1 in a similar manner. CRM1–NLP1complexes were assembled on immobilized wild-type GST–Ran, and increasing concentrations of an FG-rich C-terminal

fragment of Nup214 that is known to interact with CRM1(Fornerod et al., 1996) were added. Clearly, the Nup214 fragmentprevented NLP1 from binding to the GST–Ran–CRM1 complex

Fig. 2. NLP1 interacts with CRM1 and export cargos in a RanGTP-

dependent manner. (A) MBP–NLP1 or MBP were immobilized on beads

and incubated in the presence or absence of CRM1, NES peptide, GST–SPN1

and RanQ69L, loaded with GTP, as indicated. Note that we usually use the

RanGAP-insensitive mutant RanQ69L (Klebe et al., 1995) for this type of

experiment, although this is not required, as no RanGAP is present in the

reactions. (B) Wild-type GST–Ran loaded with GTP or GST was immobilized

on beads and incubated in the presence or absence of MBP–NLP1, CRM1 and

His–SPN1, as indicated. Bound proteins were subjected to SDS-PAGE,

followed by Coomassie Blue staining (A,B). The input in B corresponds to

10% of the material used for the binding reaction. (C) ELISA assay.

GST–Ran was loaded with either GDP or GTP, immobilized on wells of a

microtiter plate and incubated with increasing concentrations of CRM1 with

MBP–NLP1 in the presence or absence of His–SPN1, as indicated. Bound

NLP1 was detected with an anti-NLP1-antibody.

Journal of Cell Science 125 (1)146

Journ

alof

Cell

Scie

nce

(Fig. 3C), suggesting that binding of the two proteins to CRM1

is mutually exclusive. Furthermore, the Nup214 fragment

could dissociate a pre-assembled NLP1–CRM1–RanGTP complex

(data not shown). Thus, nuclear export complexes might

sequentially bind to NLP1 and Nup214 during transport, before

complex disassembly at the cytoplasmic site of the NPC by

RanBP1.

FG repeats are not crucial for the NLP1–CRM1 interaction

CRM1 has been shown to interact with a number of FG Nups

(Neville et al., 1997). For Nup214, the FG-repeat-containing C-

terminal part of the protein is required for high-affinity

interaction with CRM1 (Fornerod et al., 1997b). We therefore

analyzed the region(s) in NLP1 that are required for CRM1

binding. Various fragments and deletion mutants of NLP1 were

expressed as either MBP or GST fusions, depending on the

solubility of the proteins. The proteins were immobilized on

beads and analyzed for CRM1 binding as described above.

Fig. 4A,B summarizes our results. The details of the binding

experiments are shown in the supplementary material Fig. S2.

Surprisingly, the FG repeats of NLP1 do not seem to play a major

role in CRM1 binding. C-terminal fragments [amino acids (aa)

165–423 and 205–423], which contain the majority of the FG

repeats, showed weak binding or no binding at all. An N-terminal

fragment containing only two FG repeats (aa 1–204), by contrast,

clearly interacted with CRM1 in an SPN1- and RanGTP-

dependent manner. The putative coiled-coil region of NL

(aa 165–204) seems to contribute to CRM1-binding (compare

the full-length protein, aa 1–423 and fragment 1–423D165–204,

fragments 1–204 and 1–165 and fragments 165–423 and 205–

423). In the deletion mutant lacking the coiled-coil region, a

mutation of two FG repeats in the N-terminal part of NLP1 did

not lead to reduced CRM1 binding. Together, our biochemical

analysis shows that NLP1 can engage in interactions that are

relevant for CRM1-dependent nuclear export. Interestingly, the

FG repeats, the classic binding motif for nucleoporin–transport

receptor interactions, do not seem to be absolutely required for

CRM1 binding. We therefore generated an NLP1 mutant in

which all phenylalanine residues within FG motifs were replaced

with serine residues (NLP1FG-less). Clearly, this mutant interacted

with CRM1 in a Ran–GTP- and cargo-dependent manner, albeit

to a lesser extent than the wild-type protein (Fig. 4C).

Fig. 3. RanGAP-resistant NLP1–CRM1–RanGTP complexes are dissociated by RanBP1 and Nup214. (A) RanGAP assay. Increasing concentrations of

MBP–NLP1 (full-length) or fragments of Nup214 or Nup88 were incubated with CRM1 and [c,32P]RanGTP. GTP hydrolysis was initiated by the addition of

RanGAP. (B) MBP–NLP1 was immobilized on beads and incubated with CRM1, RanQ69L–GTP and an NES peptide. After complex formation, RanBP1 was

added for 60 minutes. (C) GST–Ran, loaded with GTP, was immobilized on beads and incubated with CRM1, MBP–NLP1 (5 mg each), an NES peptide and

increasing amounts (2.5–10 mg) of MBP–Nup214 (aa 1859–2090). Bound proteins were analyzed by SDS-PAGE and Coomassie Blue staining (B,C).

Fig. 4. FG repeats are not crucial for

NLP1–CRM1 complexes. (A) MBP- or

GST-tagged fragments of NLP1. Orange

box, zinc finger region; yellow box (C–C),

putative coiled-coil region. FG motifs are

depicted as vertical bars. Binding to

CRM1 (as analyzed in B) was

qualitatively assessed (+++, strong; ++,

intermediate; +, weak; –, no binding).

(B) NLP1 fragments were immobilized on

beads and incubated with CRM1 in the

presence or absence of His–SPN1 and

RanQ69L–GTP, as indicated. See

supplementary material Fig. S2 for entire

gels. (C) Full-length NLP1FG-less was

fused to MBP, immobilized as in B and

incubated with CRM1 in the presence or

absence of GST–SPN1 and RanQ69L–

GTP, as indicated. Binding of CRM1 was

analyzed by SDS-PAGE and Coomassie

Blue staining (B,C).

NLP1 in CRM1-dependent export 147

Journ

alof

Cell

Scie

nce

NLP1 is found at the nuclear envelope and could be mobilewithin the nucleoplasm

In a proteomic analysis, mammalian NLP1 was suggested to be a

component of the NPC, occurring at a copy number of 16(Cronshaw et al., 2002). In NLP1-overexpressing HeLa cells, theprotein was originally detected in the nucleus, being excluded

from nucleoli (Farjot et al., 1999). A myc-tagged version ofNLP1 also localized to the nuclear envelope in transfected cells(Le Rouzic et al., 2002). However, a thorough analysis of the

subcellular and subnuclear localization of the endogenous proteinhas not been performed. We therefore raised antibodies againstNLP1 in rabbits and also tested two anti-NLP1 antibodies raised

in guinea pigs (data not shown). None of these antibodies yieldedspecific and reliable signals in indirect immunofluorescenceassays, although they specifically recognized NLP1 in westernblotting. We therefore resorted to a biochemical fractionation

approach and also re-investigated the localization of NLP1using various tagged proteins in overexpressing cells. Asshown in Fig. 5A, endogenous NLP1 largely co-fractionated

with nucleoporins, as detected with the anti-FG-Nup antibodymAb414. Only small amounts of NLP1 could be detected in thecytosolic fraction. When we expressed GFP–NLP1, RFP–NLP1,

NLP1–RFP, HA–NLP1 or NLP1–HA, low expressing cellsshowed a clear signal at the nuclear envelope (Fig. 5B,arrows). Many cells also showed a clear nuclear staining,

excluding the nucleoli (Fig. 5B and data not shown). In highlyoverexpressing cells, NLP1 was also detectable in the cytoplasm.In control cells, potential binding sites for NLP1 in the NPC areprobably occupied by the endogenous protein. We therefore

expressed GFP–NLP1 in cells where the endogenous protein hadbeen depleted by specific siRNAs. The localization of GFP–NLP1 in such cells was very similar to that observed in control

cells, with some cells exhibiting a weak signal for GFP–NLP1 atthe nuclear rim and others a pan-nuclear signal (Fig. 5B, rightpanel). We next investigated whether nucleoplasmic GFP–NLP1

is a mobile protein by performing fluorescence recovery afterphotobleaching (FRAP) assays. As an example of a mobileprotein, we expressed GFP2–GST–cNLS, which is efficientlytransported into the nucleus. A small nuclear region of GFP–

NLP1- or GFP2–GST–cNLS-expressing cells was bleached andthe recovery of fluorescence was recorded. As shown in Fig. 5C,the fluorescence in the bleached area was completely restored

after ,40 seconds for GFP–NLP1 and after ,15 seconds forGFP2–GST–cNLS, suggesting that GFP–NLP1 can move withinthe nucleus, albeit with a lower mobility compared with GFP2–

GST–cNLS. This reduced mobility could be explained, forexample, by transient interactions of NLP1 with nucleic acids.Together, our results suggest that NLP1 interacts with the nuclear

envelope, in accordance with previous observations (Farjot et al.,1999; Le Rouzic et al., 2002). Depending on the expression level,it might also be found in the nuclear interior, similar to manyother nucleoporins.

NLP1 promotes CRM1-dependent nuclear export

In light of our initial results showing that the concentration

of CRM1 is rate-limiting for nuclear export (Fig. 1), we nextexpressed tagged versions of NLP1 in HeLa cells and analyzedtheir effects on nuclear transport. In cells that had been co-

transfected with an empty HA vector, our artificial reporterprotein NES–GFP2–cNLS localized mainly to the nucleus(Fig. 6A,B), similar to our results shown in Fig. 1. When cells

were co-transfected with a construct coding for NLP1–HA,

however, the percentage of cells showing a clear nuclear

localization of NES–GFP2–cNLS dropped to ,50%. Such

changes in the subcellular localization of NES–GFP2–cNLS

could result from stimulated export or inhibited import. To

distinguish between these two possibilities, we incubated

transfected cells in the presence of LMB. This treatment

resulted in a strong nuclear accumulation of NES–GFP2–cNLS,

and also occurred in cells expressing NLP1–HA (Fig. 6B).

Hence, overexpression of NLP1–HA did not inhibit nuclear

import of NES–GFP2–cNLS but rather stimulated its nuclear

export. Similar observations were made with NC2b–GFP2 as a

reporter protein and N- or C-terminally RFP-tagged NLP1.

Again, co-expression of NLP1 fusion proteins resulted in a clear

shift of NC2b–GFP2 towards the cytoplasm, compared with cells

expressing unfused RFP (Fig. 6C,D). As above, treatment of cells

Fig. 5. NLP1 is a mobile nuclear protein. (A) HeLa cells were subjected to

subcellular fractionation and proteins in the total (t), cytoplasmic (c) and

nuclear (n) fraction were analyzed by SDS-PAGE followed by western

blotting to detect NLP1 (rabbit anti-NLP1), a-tubulin and a set of

nucleoporins (monoclonal antibody 414). (B) HeLa cells that had been treated

with the specific siRNA 9 to deplete endogenous NLP1, or left untreated,

were transfected with a plasmid coding for GFP–NLP1 and analyzed by

fluorescence microscopy. Note that cells with a nuclear rim (white arrows)

can be seen under both conditions. Scale bars: 10 mm. Cell lysates were

analyzed by western blotting for levels of endogenous NLP1 and, as a loading

control, Uba2. (C) FRAP analysis of nuclear GFP–NLP1 or GFP2–GST–

cNLS. The error bars show the standard deviation from the mean recovery of

fluorescence in the bleached area of 10 cells.

Journal of Cell Science 125 (1)148

Journ

alof

Cell

Scie

nce

with LMB led to a clear nuclear accumulation of NC2b–GFP2

under all conditions, indicating stimulated nuclear export in

cells expressing RFP–NLP1 or NLP1–RFP. Next, we analyzed

the kinetics of nuclear export in control cells and in cells

overexpressing NLP1. To this end, we performed fluorescence

loss in photobleaching (FLIP) assays in cells co-expressing

NC2b–GFP2 and either RFP or RFP–NLP1. A cytoplasmic

region in RFP-positive cells was constantly bleached and the loss

of GFP fluorescence in the nuclear compartment was measured

as an indicator of the efficiency of nuclear export of NC2b–

GFP2. As shown in Fig. 6E, cells co-expressing RFP–NLP1

exhibited faster export kinetics than cells co-expressing RFP.

NLP1 is a nuclear-envelope-associated protein that could

potentially also shuttle between the nuclear envelope and the

nuclear interior. Given the stimulatory effect of NLP1 on nuclear

export in vivo, we argued that recombinant NLP1 might also

promote CRM1-dependent export in permeabilized cells. HeLa

cells that stably expressed GFP–NFAT [nuclear factor of

activated T cells (Kehlenbach et al., 1998)] were permeabilized

with digitonin and subjected to nuclear export reactions in the

presence of Ran and CRM1, with or without additional NLP1.

Clearly, NLP1 promoted nuclear export of GFP–NFAT

(supplementary material Fig. S3). This stimulation of export

was specific, as the loss of nuclear fluorescence in the presence of

NLP1 required ATP. Furthermore, nuclear export under all

conditions was inhibited by wheat germ agglutinin, a general

inhibitor of nuclear transport (data not shown), demonstrating

that the permeability barrier of the NPC was not compromised

during the reaction.

Our results, described so far, suggest that NLP1 stimulates

CRM1-dependent nuclear export under conditions where the

export receptor itself is rate-limiting. This limitation might result

from low concentrations of CRM1 and/or from comparatively

low affinities of export cargos to the receptor.

Depletion of NLP1 inhibits CRM1-dependent nuclear export

In in vitro reactions using digitonin-permeabilized cells, CRM1

and Ran were the only exogenous factors that were required

for efficient nuclear export (Kehlenbach et al., 1998). Under

these conditions, however, NLP1 is still present in nuclei of

permeabilized cells (see Fig. 5A). To specifically address the

question of whether NLP1 is required for CRM1-dependent

export, we performed siRNA experiments, reducing the cellular

concentration of the protein by RNA interference. In a similar

setup, we previously showed that depletion of nucleoporin

Nup214, but not Nup358 reduced CRM1-dependent nuclear

protein export (Hutten and Kehlenbach, 2006). As judged by

western blotting, two different siRNAs against NLP1 resulted in

a clear reduction of the NLP1 concentration, although traces of

the protein were still detectable (Fig. 7A and data not shown).

Other proteins that are involved in various steps of

nucleocytoplasmic transport were not affected by the siRNA

treatment (Fig. 7A). We first investigated nuclear export of our

reporter protein GFP–NFAT (Hutten and Kehlenbach, 2006).

Cells were treated with ionomycin to induce nuclear import of

the protein. Control cells and siRNA-treated cells showed the

same partial nuclear import of GFP–NFAT after 1 minute and

complete nuclear localization of the reporter protein after 2

Fig. 6. NLP1 promotes CRM1-dependent nuclear protein

export. (A,B) HeLa cells were co-transfected with plasmids

coding for NLP1–HA or an empty control vector (contr.) and

NES–GFP2–cNLS and treated with or without LMB, as

indicated. The subcellular localization of the reporter protein

was analyzed by microscopy (A) and quantified (B), showing

the mean percentage of cells with a predominant nuclear signal

(N.C). Error bars show the standard deviation from the mean

of three (+ LMB) or six (–LMB) independent experiments.

(C,D) HeLa cells were co-transfected with plasmids coding

for NC2b–GFP2 and RFP, RFP–NLP1 or NLP1–RFP and

treated with or without LMB, as indicated. The subcellular

localization of NC2b–GFP2 was analyzed by microscopy (C)

and quantified (D). Error bars show the standard deviation

from the mean of the percentage of cells with a predominant

nuclear signal (N.C) from three independent experiments.

(E) Cells were co-transfected with plasmids coding for NC2b–

GFP2 and either RFP or RFP–NLP1, as indicated. After

constant bleaching of a cytoplasmic region, the loss of nuclear

GFP fluorescence was analyzed by FLIP in 10 cells. Error bars

show the variation from the mean of two independent

experiments. Scale bars: 10 mm.

NLP1 in CRM1-dependent export 149

Journ

alof

Cell

Scie

nce

minutes, suggesting that nuclear import is not affected by the

depletion of NLP1 (supplementary material Fig. S4A). After a

change of medium to remove the ionomycin, nuclear export was

allowed to proceed. A clear nuclear GFP–NFAT signal was

present in ,40% of the control cells after 1 hour. In cells treated

with two different siRNAs this figure was 60–65%, indicating a

modest, yet reproducible inhibition of nuclear export under

conditions of reduced NLP1 concentrations (Fig. 7B,C). After 2

hours of export, differences between control cells and siRNA-

treated cells were less pronounced (data not shown). As an

alternative reporter for nuclear export, we used NC2b–GFP2.

This protein localizes predominantly to the nucleus (compare

Fig. 6) and obvious changes in this localization upon treatment of

cells with siRNAs against NLP1 were not expected. We therefore

resorted to a kinetic analysis of NC2b–GFP2 export by FLIP. As

shown in Fig. 7D, depletion of NLP1 by specific siRNAs led to a

small, yet reproducible reduction of the nuclear export rate

of NC2b–GFP2. Because NLP1 and its yeast homolog have

previously been implicated in RNA export (Saavedra et al., 1996;

Stutz et al., 1997; Vainberg et al., 2000), we also compared

control cells and NLP1-depleted cells by in situ hybridization,

detecting Poly(A)+ mRNA. No obvious differences in mRNA

localization were detected under our experimental conditions

(supplementary material Fig. S4B), suggesting that the observed

effects on CRM1-dependent export did not result from indirect

effects on mRNA export. In summary, reduced NLP1

concentrations resulted in decreased export of certain nuclear

proteins. NLP1, however, does not seem to be absolutely required

for transport, as export still occurred at a substantial rate upon

depletion of the protein.

DiscussionLimiting CRM1 concentrations: a means to

regulate transport?

CRM1 is an export receptor that interacts with hundreds or even

thousands of cargo proteins. Strikingly, CRM1 uses the same

hydrophobic binding pocket for interaction with different cargos,

forcing NES peptide sequences to adapt their conformation to the

rather rigid binding-site on CRM1 (Guttler et al., 2010).

Depending on the spacing of key hydrophobic residues, the

affinities of NES-containing cargos for their cognate receptor

CRM1 can vary dramatically, with low-affinity cargos prevailing

(Cook et al., 2007; Kutay and Guttinger, 2005). Functionally,

weak affinities seem to be important for efficient disassembly of

export complexes on the cytoplasmic side of the NPC (Engelsma

et al., 2004). Furthermore, they prevent cargos from binding to

CRM1 in the cytoplasm in the absence of RanGTP (Kutay and

Guttinger, 2005). The consequences of different affinities on the

subcellular localization of individual nucleocytoplasmic shuttling

proteins has not been investigated so far. Our results now show

that the cellular CRM1 concentrations are rate limiting for

nuclear export in HeLa cells, at least under conditions of

overexpression of cargo proteins. Elevated CRM1 levels have

been observed in many tumor cells including cervical cancers

(Noske et al., 2008; van der Watt et al., 2009) and could also be

effective in our HeLa cells. Hence, export of a subset of

endogenous proteins in primary cells could well be limited under

certain conditions by the available pool of the export receptor,

allowing the cells to regulate export simply by adjusting the

CRM1 concentration. Under conditions where many NES-

containing proteins compete for binding sites on CRM1, the

formation of export complexes containing low-affinity CRM1

substrates could also be promoted by accessory factors. One such

factor is the Ran-binding protein RanBP3, which enhances the

affinity of CRM1 for cargo proteins and for RanGTP. As a

consequence, RanBP3 stimulates CRM1-dependent export in

permeabilized cells (Englmeier et al., 2001; Lindsay et al., 2001).

Another cofactor in CRM1-dependent export is the nucleoporin

Nup98, which also interacts with the export receptor in a

RanGTP-dependent manner (Oka et al., 2010). Here, the authors

used antibody-injection experiments and suggested a role of

Nup98 in export of overexpressed transport cargos in living cells.

Fig. 7. Depletion of NLP1 inhibits CRM1-

dependent protein export. (A) HeLa cells were

treated with siRNA 8 against NLP1 and cell lysates

were analyzed by SDS-PAGE and western blotting.

(B,C) HeLa NFAT cells were treated with siRNA 8

or 9 against NLP1 and nuclear import of GFP–

NFAT was induced with ionomycin. After washing

the cells to remove the ionomycin, nuclear export

was analyzed immediately (0 h) or after 1 hour,

by microscopy Scale bar: 10 mm.

(C) Quantification of cells showing a predominant

nuclear localization of GFP–NFAT after 1 hour.

Error bars show the standard deviation from the

mean of three independent experiments. (D) HeLa

cells were treated with siRNA 9 against NLP1 and

transfected with a plasmid coding for NC2b–GFP2.

Nuclear export of NC2b–GFP2 was analyzed by

FLIP, bleaching a cytoplasmic region and recording

the loss of nuclear GFP fluorescence over time.

Error bars show the standard deviation from the

mean of four independent experiments.

Journal of Cell Science 125 (1)150

Journ

alof

Cell

Scie

nce

In our study, we analyzed the role of a nucleoporin-like protein,NLP1, in CRM1-mediated export. NLP1 had previously been

implicated in nuclear export of certain mRNAs (Katahira et al.,1999; Kendirgi et al., 2005). Although it had been shown tointeract with CRM1 in two-hybrid assays, its function in protein

export in mammalian cells has remained unclear. We now showthat NLP1 promotes CRM1-dependent export of certain cargos invitro and in vivo. Together with other accessory factors, NLP1might contribute to the control of nuclear export of distinct cargo

proteins.

NLP1 promotes CRM1-dependent export

NLP1 was originally so named because of the characteristic FGrepeats, as they also occur in certain nucleoporins. Indeed, the

protein was identified as a component of the NPC in a proteomicanalysis (Cronshaw et al., 2002). In one study, the yeast homologueRip1p/Nup42p has been found to reside in the nucleus and at the

nuclear and the cytoplasmic face of the NPC (Strahm et al., 1999)and in another study exclusively on the cytoplasmic sideof the NPC (Rout et al., 2000). Our results now suggest thatmammalian NLP1 is associated with the nuclear envelope and could

also be mobile within the nucleus, similar to Nup98. The levels ofintranuclear NLP1 could well depend on the expression level of theNLP1 gene. To unequivocally demonstrate the existence of an

intranuclear pool of NLP1 and further analyze its significance,specific antibodies that are suitable for immunofluorescence orimmunoelectron microscopy will be required.

Overexpression as well as siRNA depletion experiments pointto a role of NLP1 in CRM1-mediated nuclear protein export. Theresults of the overexpression experiments can be easily explained

if we assume that a nuclear pool of NLP1 does exist underphysiological conditions and could thus support complexformation as discussed below. If a nuclear pool only resulted

from overexpression of the protein, we would have to assume thatthe levels of endogenous NLP1 at the nuclear envelope are notsaturated in all cells. Upon overexpression, additional bindingsites at the nuclear envelope and/or nuclear pore are then

occupied by NLP1, resulting in stimulation of CRM1-dependentexport. However, the siRNA depletion experiments in whichwe observed a small reduction of CRM1-dependent transport,

suggest that NLP1 is not absolutely required for nuclear export.Instead, it functions as an accessory factor, e.g. for proteins thaton their own have a low affinity for the export receptor. In cells

that do not overexpress transport receptors as many cancer cellsdo (Noske et al., 2008; van der Watt et al., 2009), limiting CRM1concentrations might prevail. It will be interesting to investigate

the functional relevance of NLP1 in such cells in the future.

In binding assays, we could demonstrate that: (1) NES cargosas well as RanGTP enhanced binding of CRM1 to NLP1; (2)

NLP1 enhanced binding of cargo to CRM1–RanGTP complexes;and (3) CRM1 (indirectly) enhanced binding of NLP1 toRanGTP. These results suggest that distinct trimeric and

tetrameric export complexes can assemble in the nucleus.Strikingly, the FG repeats of NLP1 do not play a major role inCRM1 binding, although an FG-rich region of Nup214 seems tocompete for the same binding site on CRM1. It will be interesting

to compare the interaction mode of CRM1 with the two proteinsin detail.

High-affinity NES cargos might form trimeric exportcomplexes with RanGTP and CRM1 and be exported assuch. Alternatively, a trimeric ‘pre-export complex’, containing

CRM1, RanGTP and NLP1 might initially form in the nucleus oron the nuclear side of the NPC and accommodate a fourth

component, a high- or low-affinity NES cargo. In such trimericand tetrameric complexes, Ran is resistant to RanGAP-induced

GTP hydrolysis. RanGAP, which promotes GTP hydrolysis bymore than 1000-fold (Bischoff et al., 1994) is restricted to thecytoplasm but could occasionally enter the nucleus, leading to

deleterious effects on the Ran gradient. In that sense, NLP1would help to protect nuclear Ran from GTP hydrolysis.

Furthermore, it would compensate for low affinities of NESs,which on their own are not sufficient for tight CRM1 binding,

at least not under conditions where high-affinity cargos arecompeting for the same binding site. How could such amechanism facilitate export of low-affinity cargos, when also

high-affinity CRM1 substrates, such as SPN1, would profit fromenhanced export complex formation? In the simplest scenario,

certain NES cargos could directly interact with NLP1, prior toexport complex formation, resulting in a stabilization of thetetrameric (cargo–NLP1–CRM1–RanGTP) complex. Indeed, our

initial results suggest that the HIV-1 Rev protein, a CRM1 cargoof comparatively low affinity (Paraskeva et al., 1999), can

directly interact with NLP1, independent of CRM1 or RanGTP(data not shown). In such a complex, the NES of the cargo

protein might be exposed upon NLP1 binding, promotingsubsequent CRM1 interaction. Alternatively, NLP1 could levelthe differences in affinities as observed in direct cargo–CRM1

interactions. In that case, increasing the cellular concentrationof NLP1 (or regulating its activity by post-translational

modifications) would have a more profound effect on low-affinity CRM1 cargos than on high-affinity substrates. A similar

observation was made for RanBP3, which increased the bindingof a GST–NES substrate to CRM1 while decreasing binding ofSPN1 (Englmeier et al., 2001).

After complex formation in the nucleus, the FG-rich C-terminal region of NLP1, which is not required for CRM1

binding, could facilitate translocation of the export complexthrough the transport channel of the NPC. Subsequently,disassembly of export complexes is the terminal step of

transport on the cytoplasmic side of the NPC. In the simplestmodel, this is accomplished by the concerted action of RanBP1 or

the RanGTP binding domains of the cytoplasmic nucleoporinNup358 (RanBP2) and the GTPase-activating protein RanGAP(Kehlenbach et al., 1999). For export complexes containing

NLP1 as an integral component, an additional factor could comeinto play. Nup214, a bona fide FG nucleoporin that also localizes

to the cytoplasmic side of the NPC, binds to CRM1–RanGTP–NES complexes with high affinity (Hutten and Kehlenbach,

2006). Binding of NLP1 and Nup214 to such complexes ismutually exclusive (Fig. 3C), and Nup214 could replace NLP1 inthe complex. It will be interesting to compare the interaction of

CRM1 with the two proteins at the structural level, because FGrepeats appear to be important for Nup214 binding but not for

NLP1 binding. The export complex, now tethered to Nup214,would be well positioned for disassembly by RanBP1 and

RanGAP, a large portion of which stably associates with Nup358(Mahajan et al., 1997; Matunis et al., 1996). Nup358 alsocontains four Ran-binding domains that are functionally

equivalent to that of RanBP1. Hence, soluble cytoplasmicfactors would be dispensable for export complex dissociation,

in line with previous observations (Kehlenbach et al., 1999).Interestingly, depletion of NLP1 reduced CRM1-dependent

NLP1 in CRM1-dependent export 151

Journ

alof

Cell

Scie

nce

export of NC2b and NFAT, two proteins that are also affected by

the depletion of Nup214 (Stephanie Roloff and R.H.K.,unpublished observations) (Hutten and Kehlenbach, 2006).Together, our data suggest a new function for NLP1 inpromoting CRM1-dependent nuclear export of a subset of proteins.

Materials and MethodsCell culture and transfections

HeLa P4 cells (Charneau et al., 1994) and HeLa NFAT cells (Hutten andKehlenbach, 2006) were grown in DMEM (Gibco) containing 4.5 g/l glucose, 10%FCS, 2 mM glutamine, 100 U/ml penicillin and 100 mg/ml streptomycin. Cellswere transfected using the calcium phosphate method (Ausubel et al., 1994) orPolyfect transfection reagent (Qiagen) according to the instructions of themanufacturer and cultured for 48 hours.

Plasmids

Constructs coding for CRM1–HA (Hilliard et al., 2010), NC2b–GFP2 (Kahle et al.,2009), NES–GFP2–cNLS [Rev(47–116)–GFP2–cNLS] and Rev68–90–GFP2–M9core (Hutten et al., 2008; Hutten et al., 2009) have been describedpreviously. Human NLP1 was PCR amplified from cDNA and cloned intopMal-C2 using BamHI and HindIII, generating MBP–NLP1. NLP1 fragments werePCR amplified from MBP–NLP1 and cloned into pMal-C2 or pGEX-KG usingBamHI and HindIII. Deletion mutants (MBP–NLP1D165–204) and point mutants(MBP–NLP1 FGFG14,15,95,96AAAA, D165–204) were generated by site-directedmutagenesis. A mutant NLP1 (NLP1FG-less) with all phenylalanine residues withinthe FG motifs replaced by serine residues was synthesized by GeneArt(Regensburg, Germany). For construction of GFP–NLP1, RFP–NLP1, NLP1–RFP and NLP1–HA, NLP1 sequences were amplified from MBP–NLP1 andcloned into pEGFP-C1, pmRFP-C1, pmRFP-N1 (Clontech) or pcDNA3.1(+),containing an HA tag. The plasmid coding for the FRAP reporter GFP2–GST–cNLS was kindly provided by Sonja Neimanis (University of Gottingen,Germany). For generation of His–SPN1 and GFP–SPN1, SPN1 sequences werePCR-amplified from GST–SPN1 (Strasser et al., 2004) and cloned into pET30b orpEGFP-C1, respectively. Constructs were verified by DNA sequencing.

Protein expression and purification

MBP–NLP1 and GST–NLP1 fusions were expressed in Escherichia coli

BL21(DE3) cells upon induction with 500 mM isopropyl-b-D-thiogalactopyranosid (IPTG) for 3.5 hours at 25 C. Cells were lysed in PBS,2 mM EDTA, 1% Triton X-100, 5 mM dithiothreitol (DTT), 1 mg/ml each ofaprotinin, leupeptin and pepstatin, and centrifuged for 30 minutes at 100,000 g.After diluting the supernatant 1:5 with PBS, 10 mM MgCl2, 5 mM DTT, 1 mg/mleach of aprotinin, leupeptin and pepstatin, proteins were purified by affinitychromatography using amylose–resin beads (New England Biolabs) for MBP-tagged fragments or glutathione–Sepharose (GE Healthcare) for GST fusionproteins. After several washing steps, proteins were eluted with either 20 mMglutathione or 10 mM maltose in 10 mM Hepes pH 8.0, 5% glycerol, 150 mMNaCl, 1 mM DTT and 1 mg/ml of protease inhibitors. His–NLP1 (1–204) wasexpressed as described for MBP–NLP1. For purification, a bacterial pellet waslysed in 50 mM Tris pH 8.0, 2% Triton X-100, 2 mM DTT and 1 mg/ml eachof aprotinin, leupeptin and pepstatin. After centrifugation for 20 minutes at100,000 g, the pellet was resuspended in 50 mM Tris pH 8.0, 4 M urea, 1 mMDTT and protease inhibitors and centrifuged as above. The resulting pellet wassolubilized in binding buffer (50 mM Tris pH 8.0, 8 M urea, 1 mM DTT andprotease inhibitors). After centrifugation as above, the supernatant was added toNi-NTA agarose beads (Qiagen) and incubated for 1 hour. After several washingsteps, the protein was eluted with binding buffer containing 300 mM imidazoleand dialyzed against PBS. Precipitated protein was used for antibody production.GST–Ran and His–SPN1 were expressed in BL21(DE3) upon induction with500 mM isopropyl b-D-1-thiogalactopyranoside and purified according to standardprotocols. GST–SPN1 (Strasser et al., 2004), RanBP1 (Kehlenbach et al., 2001),RanGAP (Mahajan et al., 1997), CRM1–His (Guan et al., 2000), Ran andRanQ69L (Melchior et al., 1995), MBP–Nup88 (aa 500-741) and MBP–Nup214(aa1859–2090) (Hutten and Kehlenbach, 2006) were purified as described before.His–Nup214 was expressed in BL21(DE3) and purified as described above forMBP–NLP1 using Ni-NTA agarose beads and 300 mM imidazole for proteinelution. All proteins were dialyzed against transport buffer (TPB; 20 mM Hepes–KOH, pH 7.3, 110 mM potassium acetate, 2 mM magnesium acetate, 1 mMEGTA, 2 mM DTT and 1 mg/ml each of aprotinin, leupeptin, pepstatin), frozen inliquid nitrogen and stored at –80 C. Ran was loaded with GTP, GDP or[c-32P]GTP as described previously (Kehlenbach et al., 1999).

RNA interference

Cells were transfected with small interfering RNAs (siRNAs) from Ambion (NLP1siRNA8, 59-agguaauaauagacguggatt-39, corresponding to nucleotides 120–138 andNLP1 siRNA9, 59-gagcuucaacuaacaggaatt-39, corresponding to nucleotides 257–275)

at a final concentration of 100 nM using Oligofectamine (Invitrogen), according to theinstructions of the manufacturer. Control cells were treated with transfection reagentonly. After 24 hours of siRNA treatment, cells were transfected with reporter proteinconstructs of interest and analyzed after 20–24 hours.

Antibodies

Antibodies against NLP1 were raised in rabbits by injecting His–NLP1 1–204 andaffinity purified using MBP–NLP1 coupled to CNBr beads. For the detectionof HA-epitope-tagged proteins by immunofluorescence, a monoclonal mouseanti-HA antibody (16B12, Covance) was used. The rabbit anti-Nup214 antibody(Hutten and Kehlenbach, 2006) and the goat anti-RanGAP antibody (Hutten et al.,2008; Pichler et al., 2002) were described previously. The mouse anti-lamin A/C,anti-Ran, anti-RanBP1 and anti-RanBP3 antibodies were obtained from BDBioscience, the mouse anti-a-tubulin antibody was obtained from Sigma. Theanti-CRM1 antibody was raised in goat against a C-terminal decapeptide asdescribed before (Kehlenbach et al., 1998). The goat-anti Uba2 antibody was akind gift from Frauke Melchior (University of Heidelberg, Germany). Forimmunofluorescence, donkey anti-mouse Alexa Fluor 594 and donkey anti-rabbitAlexa Fluor 594 (1:1000; Molecular Probes) were used as secondary antibodies.For immunoblotting, HRP-coupled donkey anti-goat, donkey anti-mouse ordonkey anti-rabbit IgG (1:5000; Dianova) were used as secondary antibodies.

Indirect immunofluorescence and microscopy

Immunofluorescence staining was performed as described before (Hutten andKehlenbach, 2006). Cells were analyzed using a Zeiss Axioskop 2 microscope witha 1006 Plan-Neofluar 1.3 NA water-corrected objective and appropriate filtersettings and processed using AxioVision Rel. 4.8 LE and Adobe Photoshop 6.0.For quantification of subcellular localization of reporter proteins, cells weregrouped into two categories: N.C (majority of the reporter protein in the nucleus)and others (either N5C; equal distribution between nucleus and cytoplasm orC.N; predominant cytoplasmic localization). Quantification was performed fromat least three independent experiments, counting more than 100 cells with similarexpression levels. Statistical significance of the data was analyzed by a two-tailed,heteroscedastic Student’s t-test. P-values ,0.05 were considered as statisticallysignificant.

Live cell imaging

FRAP and FLIP analyses were performed at 37 C with a LSM 510 Meta confocalmicroscope (Zeiss) using a LCI Plan-Neofluar 636 1.3 NA water-correctedobjective in a temperature-controlled chamber. Cells were grown on LabTec-chambers (Nunc) and transferred to CO2-free medium (Invitrogen) before theanalysis. The mobility of GFP–NLP1 was analyzed by FRAP. After 20 scans atlow laser intensity (1.2% 488-nm laser transmission), an area of 10630 pixels(1.3564.15 mm) in the nucleus was bleached (100% transmission of the 488-nmlaser, 40 iterations). Subsequently, scanning of the bleached area was continued atlow laser intensity for 380 pictures (154 mseconds/picture). Data analysis waspreformed with LSM software (Zeiss) and Excel (Microsoft), normalizing theoriginal fluorescence in the bleached area to one. FLIP analysis of nuclear exportwas performed essentially as described before (Hilliard et al., 2010).

Nuclear transport in HeLa-cells expressing GFP–NFAT

Nuclear import of GFP–NFAT (Hutten and Kehlenbach, 2006) was induced by theaddition of 300 nM ionomycin (Sigma) and cells were incubated at 37 C forvarious periods of time, fixes with 4% formaldehyde and analyzed by fluorescencemicroscopy. Export of GFP–NFAT was analyzed as described previously (Huttenand Kehlenbach, 2006).

Binding studies

MBP or GST fusion proteins (5 mg) were immobilized on 20 ml amylose–agaroseor glutathione–agarose beads that had been preincubated with 10 mg/ml BSA inTPB. The beads were incubated with 5 mg of each protein of interest or 10 mMNES peptide [NS2 protein of minute virus of mice, CVDEMTKKFGTLTIHDTEK(Askjaer et al., 1999)] in 300 ml TPB containing 2 mg/ml BSA. After 75 minutesat 4 C, beads were washed three times with TPB. Bound proteins were eluted withSDS sample buffer and subjected to SDS-PAGE, followed by Coomassie Bluestaining or western blotting.

ELISA

A 96-well plate (Immuno 96 MicroWellTM Solid Plates; Nunc) was coatedovernight at 4 C with 300 ng GST–Ran per well loaded with GDP or GTP in 50 mlTPB. The wells were blocked with 3% BSA in TPB for 30 minutes at roomtemperature and 300 ng MBP–NLP1 and increasing concentrations of CRM1–His,with or without 300 ng His–SPN1, were added. After incubation for 75 minutes atroom temperature and several washing steps, bound NLP1 was detected using therabbit anti-NLP1 antibody (1:5000 in TPB + 3% BSA) and HRP-conjugateddonkey anti-rabbit IgG (1:2000 in TPB + 3% BSA) as secondary antibody. For thecolorimetric reaction, the TMB substrate reagent set from Millipore was used. The

Journal of Cell Science 125 (1)152

Journ

alof

Cell

Scie

nce

absorbance was measured at 420 nm using a plate reader (Appliskan, ThermoElectron Corporation).

Cell fractionation

HeLa cells were trypsinized, washed with PBS and permeabilized with 0.01%

digitonin in PBS with 1 mg/ml each of aprotinin, leupeptin and pepstatin. Nucleiwere separated from cytosol by centrifugation at 1000 g for 5 minutes and washed

twice with PBS. Fractions corresponding to 100,000 cells were analyzed by SDS-PAGE, followed by western blotting.

SDS-PAGE and western blotting

Proteins were analyzed by SDS-PAGE and western blotting using standardmethods. The ECL system (Millipore) was used for visualization of proteins.

RanGAP-assays

RanGAP assays were performed as described previously (Askjaer et al., 1999;

Kehlenbach et al., 2001), with 16 nM RanGAP and increasing concentrations ofMBP–NLP1, MBP–Nup214 (aa 1859–2090) or MBP–Nup88 (aa 500–741).

In situ hybridization

In situ hybridization was performed as described previously (Hutten and

Kehlenbach, 2006).

AcknowledgementsWe thank Frauke Melchior, Detlef Doenecke, Ralf Ficner and VolkerCordes for valuable reagents, Saskia Hutten for preparation of Nup88fragments and Blanche Schwappach and Ruth Geiss-Friedlander forhelpful comments on the manuscript.

FundingThe project was partially funded by DeutscheForschungsgemeinschaft [grant number KE 660/9-1 to R.H.K.].

Supplementary material available online at

http://jcs.biologists.org/lookup/suppl/doi:10.1242/jcs.090316/-/DC1

ReferencesAskjaer, P., Bachi, A., Wilm, M., Bischoff, F. R., Weeks, D. L., Ogniewski, V.,

Ohno, M., Niehrs, C., Kjems, J., Mattaj, I. W. et al. (1999). RanGTP-regulated

interactions of CRM1 with nucleoporins and a shuttling DEAD-box helicase. Mol.

Cell. Biol. 19, 6276-6285.

Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith,

J. A. and Struhl, K. (1994). Current Protocols in Molecular Biology. New York:

Greene Publishing Associates and Wiley-Interscience.

Bernad, R., van der Velde, H., Fornerod, M. and Pickersgill, H. (2004). Nup358/

RanBP2 attaches to the nuclear pore complex via association with Nup88 and

Nup214/CAN and plays a supporting role in CRM1-mediated nuclear protein export.

Mol. Cell. Biol. 24, 2373-2384.

Bernad, R., Engelsma, D., Sanderson, H., Pickersgill, H. and Fornerod, M. (2006).

Nup214-Nup88 nucleoporin subcomplex is required for CRM1-mediated 60 S

preribosomal nuclear export. J. Biol. Chem. 281, 19378-19386.

Bischoff, F. R., Klebe, C., Kretschmer, J., Wittinghofer, A. and Ponstingl, H. (1994).

RanGAP1 induces GTPase activity of nuclear Ras-related Ran. Proc. Natl. Acad. Sci.

USA 91, 2587-2591.

Chadrin, A., Hess, B., San Roman, M., Gatti, X., Lombard, B., Loew, D., Barral, Y.,

Palancade, B. and Doye, V. (2010). Pom33, a novel transmembrane nucleoporin

required for proper nuclear pore complex distribution. J. Cell Biol. 189, 795-811.

Charneau, P., Mirambeau, G., Roux, P., Paulous, S., Buc, H. and Clavel, F. (1994).

HIV-1 reverse transcription. A termination step at the center of the genome. J. Mol.

Biol. 241, 651-662.

Cook, A., Bono, F., Jinek, M. and Conti, E. (2007). Structural biology of

nucleocytoplasmic transport. Annu. Rev. Biochem. 76, 647-671.

Cronshaw, J. M., Krutchinsky, A. N., Zhang, W., Chait, B. T. and Matunis, M. J.

(2002). Proteomic analysis of the mammalian nuclear pore complex. J. Cell Biol. 158,

915-927.

Dong, X., Biswas, A., Suel, K. E., Jackson, L. K., Martinez, R., Gu, H. and Chook,

Y. M. (2009). Structural basis for leucine-rich nuclear export signal recognition by

CRM1. Nature 458, 1136-1141.

Engelsma, D., Bernad, R., Calafat, J. and Fornerod, M. (2004). Supraphysiological

nuclear export signals bind CRM1 independently of RanGTP and arrest at Nup358.

EMBO J. 23, 3643-3652.

Englmeier, L., Fornerod, M., Bischoff, F. R., Petosa, C., Mattaj, I. W. and Kutay, U.

(2001). RanBP3 influences interactions between CRM1 and its nuclear protein export

substrates. EMBO Rep. 2, 926-932.

Farjot, G., Sergeant, A. and Mikaelian, I. (1999). A new nucleoporin-like proteininteracts with both HIV-1 Rev nuclear export signal and CRM-1. J. Biol. Chem. 274,

17309-17317.

Fischer, U., Huber, J., Boelens, W. C., Mattaj, I. W. and Luhrmann, R. (1995). TheHIV-1 Rev activation domain is a nuclear export signal that accesses an export

pathway used by specific cellular RNAs. Cell 82, 475-483.

Floer, M. and Blobel, G. (1999). Putative reaction intermediates in Crm1-mediatednuclear protein export. J. Biol. Chem. 274, 16279-16286.

Fornerod, M., Boer, J., van Baal, S., Morreau, H. and Grosveld, G. (1996).Interaction of cellular proteins with the leukemia specific fusion proteins DEK-CAN

and SET-CAN and their normal counterpart, the nucleoporin CAN. Oncogene 13,1801-1808.

Fornerod, M., Ohno, M., Yoshida, M. and Mattaj, I. W. (1997a). CRM1 is an export

receptor for leucine-rich nuclear export signals. Cell 90, 1051-1060.

Fornerod, M., van Deursen, J., van Baal, S., Reynolds, A., Davis, D., Murti, K. G.,

Fransen, J. and Grosveld, G. (1997b). The human homologue of yeast CRM1 is in a

dynamic subcomplex with CAN/Nup214 and a novel nuclear pore component Nup88.EMBO J. 16, 807-816.

Fried, H. and Kutay, U. (2003). Nucleocytoplasmic transport: taking an inventory.

Cell. Mol. Life Sci. 60, 1659-1688.

Fukuda, M., Asano, S., Nakamura, T., Adachi, M., Yoshida, M., Yanagida, M. and

Nishida, E. (1997). CRM1 is responsible for intracellular transport mediated by the

nuclear export signal. Nature 390, 308-311.

Gerace, L., Ottaviano, Y. and Kondor-Koch, C. (1982). Identification of a majorpolypeptide of the nuclear pore complex. J. Cell Biol. 95, 826-837.

Goldfarb, D. S. (2009). At loose ends: a guide to the translocation channel of the nuclearpore complex for dummies. In Nuclear Transport (ed. R. H. Kehlenbach), pp 54-68.

Austin, Texas: Landes Bioscience.

Guan, T., Kehlenbach, R. H., Schirmer, E. C., Kehlenbach, A., Fan, F., Clurman,

B. E., Arnheim, N. and Gerace, L. (2000). Nup50, a nucleoplasmically oriented

nucleoporin with a role in nuclear protein export. Mol. Cell Biol. 20, 5619-5630.

Guttler, T., Madl, T., Neumann, P., Deichsel, D., Corsini, L., Monecke, T., Ficner,

R., Sattler, M. and Gorlich, D. (2010). NES consensus redefined by structures of

PKI-type and Rev-type nuclear export signals bound to CRM1. Nat. Struct. Mol. Biol.

17, 1367-1376.

Hallberg, E., Wozniak, R. W. and Blobel, G. (1993). An integral membrane protein of

the pore membrane domain of the nuclear envelope contains a nucleoporin-likeregion. J. Cell Biol. 122, 513-521.

Hilliard, M., Frohnert, C., Spillner, C., Marcone, S., Nath, A., Lampe, T.,

Fitzgerald, D. J. and Kehlenbach, R. H. (2010). The anti-inflammatoryprostaglandin 15-deoxy-delta(12,14)-PGJ2 inhibits CRM1-dependent nuclear proteinexport. J. Biol. Chem. 285, 22202-22210.

Hutten, S. and Kehlenbach, R. H. (2006). Nup214 is required for CRM1-dependentnuclear protein export in vivo. Mol. Cell. Biol. 26, 6772-6785.

Hutten, S. and Kehlenbach, R. H. (2007). CRM1-mediated nuclear export: to the poreand beyond. Trends Cell Biol. 17, 193-201.

Hutten, S., Flotho, A., Melchior, F. and Kehlenbach, R. H. (2008). The Nup358-

RanGAP complex is required for efficient importin alpha/beta-dependent nuclearimport. Mol. Biol. Cell 19, 2300-2310.

Hutten, S., Walde, S., Spillner, C., Hauber, J. and Kehlenbach, R. H. (2009). The

nuclear pore component Nup358 promotes transportin-dependent nuclear import. J.

Cell Sci. 122, 1100-1110.

Kahle, J., Piaia, E., Neimanis, S., Meisterernst, M. and Doenecke, D. (2009).

Regulation of nuclear import and export of negative cofactor 2. J. Biol. Chem. 284,9382-9393.

Katahira, J., Strasser, K., Podtelejnikov, A., Mann, M., Jung, J. U. and Hurt, E.

(1999). The Mex67p-mediated nuclear mRNA export pathway is conserved fromyeast to human. EMBO J. 18, 2593-2609.

Kehlenbach, R. H., Dickmanns, A. and Gerace, L. (1998). Nucleocytoplasmic

shuttling factors including Ran and CRM1 mediate nuclear export of NFAT in vitro.J. Cell Biol. 141, 863-874.

Kehlenbach, R. H., Dickmanns, A., Kehlenbach, A., Guan, T. and Gerace, L. (1999).A role for RanBP1 in the release of CRM1 from the nuclear pore complex in aterminal step of nuclear export. J. Cell Biol. 145, 645-657.

Kehlenbach, R. H., Assheuer, R., Kehlenbach, A., Becker, J. and Gerace, L. (2001).Stimulation of nuclear export and inhibition of nuclear import by a Ran mutantdeficient in binding to Ran-binding protein 1. J. Biol. Chem. 276, 14524-14531.

Kendirgi, F., Rexer, D. J., Alcazar-Roman, A. R., Onishko, H. M. and Wente, S. R.

(2005). Interaction between the shuttling mRNA export factor Gle1 and thenucleoporin hCG1: a conserved mechanism in the export of Hsp70 mRNA. Mol.

Biol. Cell 16, 4304-4315.

Klebe, C., Bischoff, F. R., Ponstingl, H. and Wittinghofer, A. (1995). Interaction ofthe nuclear GTP-binding protein Ran with its regulatory proteins RCC1 andRanGAP1. Biochemistry 34, 639-647.

Kutay, U. and Guttinger, S. (2005). Leucine-rich nuclear-export signals: born to beweak. Trends Cell Biol. 15, 121-124.

Le Rouzic, E., Mousnier, A., Rustum, C., Stutz, F., Hallberg, E., Dargemont, C. and

Benichou, S. (2002). Docking of HIV-1 Vpr to the nuclear envelope is mediated bythe interaction with the nucleoporin hCG1. J. Biol. Chem. 277, 45091-45098.

Lindsay, M. E., Holaska, J. M., Welch, K., Paschal, B. M. and Macara, I. G. (2001).Ran-binding protein 3 is a cofactor for Crm1-mediated nuclear protein export. J. Cell

Biol. 153, 1391-1402.

NLP1 in CRM1-dependent export 153

Journ

alof

Cell

Scie

nce

Mahajan, R., Delphin, C., Guan, T., Gerace, L. and Melchior, F. (1997). A smallubiquitin-related polypeptide involved in targeting RanGAP1 to nuclear pore complexprotein RanBP2. Cell 88, 97-107.

Mansfeld, J., Guttinger, S., Hawryluk-Gara, L. A., Pante, N., Mall, M., Galy, V.,

Haselmann, U., Muhlhausser, P., Wozniak, R. W., Mattaj, I. W. et al. (2006). Theconserved transmembrane nucleoporin NDC1 is required for nuclear pore complexassembly in vertebrate cells. Mol. Cell 22, 93-103.

Matunis, M. J., Coutavas, E. and Blobel, G. (1996). A novel ubiquitin-likemodification modulates the partitioning of the Ran-GTPase-activating proteinRanGAP1 between the cytosol and the nuclear pore complex. J. Cell Biol. 135,1457-1470.

Melchior, F., Sweet, D. J. and Gerace, L. (1995). Analysis of Ran/TC4 function innuclear protein import. Methods Enzymol. 257, 279-291.

Monecke, T., Guttler, T., Neumann, P., Dickmanns, A., Gorlich, D. and Ficner, R.

(2009). Crystal structure of the nuclear export receptor CRM1 in complex withSnurportin1 and RanGTP. Science 324, 1087-1091.

Neville, M., Stutz, F., Lee, L., Davis, L. I. and Rosbash, M. (1997). The importin-betafamily member Crm1p bridges the interaction between Rev and the nuclear porecomplex during nuclear export. Curr. Biol. 7, 767-775.

Noske, A., Weichert, W., Niesporek, S., Roske, A., Buckendahl, A. C., Koch, I.,Sehouli, J., Dietel, M. and Denkert, C. (2008). Expression of the nuclear exportprotein chromosomal region maintenance/exportin 1/Xpo1 is a prognostic factor inhuman ovarian cancer. Cancer 112, 1733-1743.

Oka, M., Asally, M., Yasuda, Y., Ogawa, Y., Tachibana, T. and Yoneda, Y. (2010).The mobile FG nucleoporin Nup98 is a cofactor for Crm1-dependent protein export.Mol. Biol. Cell 21, 1885-1896.

Ossareh-Nazari, B., Bachelerie, F. and Dargemont, C. (1997). Evidence for a role ofCRM1 in signal-mediated nuclear protein export. Science 278, 141-144.

Paraskeva, E., Izaurralde, E., Bischoff, F. R., Huber, J., Kutay, U., Hartmann, E.,

Luhrmann, R. and Gorlich, D. (1999). CRM1-mediated recycling of snurportin 1 tothe cytoplasm. J. Cell Biol. 145, 255-264.

Pichler, A., Gast, A., Seeler, J. S., Dejean, A. and Melchior, F. (2002). Thenucleoporin RanBP2 has SUMO1 E3 ligase activity. Cell 108, 109-120.

Rout, M. P., Aitchison, J. D., Suprapto, A., Hjertaas, K., Zhao, Y. and Chait, B. T.

(2000). The yeast nuclear pore complex: composition, architecture, and transportmechanism. J. Cell Biol. 148, 635-651.

Saavedra, C., Tung, K. S., Amberg, D. C., Hopper, A. K. and Cole, C. N. (1996).Regulation of mRNA export in response to stress in Saccharomyces cerevisiae. Genes

Dev. 10, 1608-1620.Saavedra, C. A., Hammell, C. M., Heath, C. V. and Cole, C. N. (1997). Yeast heat

shock mRNAs are exported through a distinct pathway defined by Rip1p. Genes Dev.

11, 2845-2856.Stade, K., Ford, C. S., Guthrie, C. and Weis, K. (1997). Exportin 1 (Crm1p) is an

essential nuclear export factor. Cell 90, 1041-1050.

Stavru, F., Hulsmann, B. B., Spang, A., Hartmann, E., Cordes, V. C. and Gorlich, D.

(2006). NDC1: a crucial membrane-integral nucleoporin of metazoan nuclear pore

complexes. J. Cell Biol. 173, 509-519.

Strahm, Y., Fahrenkrog, B., Zenklusen, D., Rychner, E., Kantor, J., Rosbach, M.

and Stutz, F. (1999). The RNA export factor Gle1p is located on the cytoplasmicfibrils of the NPC and physically interacts with the FG-nucleoporin Rip1p, the

DEAD-box protein Rat8p/Dbp5p and a new protein Ymr 255p. EMBO J. 18, 5761-5777.

Strasser, A., Dickmanns, A., Schmidt, U., Penka, E., Urlaub, H., Sekine, M.,

Luhrmann, R. and Ficner, R. (2004). Purification, crystallization and preliminarycrystallographic data of the m3G cap-binding domain of human snRNP import factorsnurportin 1. Acta Crystallogr. D Biol. Crystallogr. 60, 1628-1631.

Stutz, F., Neville, M. and Rosbash, M. (1995). Identification of a novel nuclear pore-

associated protein as a functional target of the HIV-1 Rev protein in yeast. Cell 82,495-506.

Stutz, F., Kantor, J., Zhang, D., McCarthy, T., Neville, M. and Rosbash, M. (1997).

The yeast nucleoporin rip1p contributes to multiple export pathways with no essentialrole for its FG-repeat region. Genes Dev. 11, 2857-2868.

Terry, L. J. and Wente, S. R. (2009). Flexible gates: dynamic topologies and functions

for FG nucleoporins in nucleocytoplasmic transport. Eukaryot. Cell 8, 1814-1827.

Vainberg, I. E., Dower, K. and Rosbash, M. (2000). Nuclear export of heat shock andnon-heat-shock mRNA occurs via similar pathways. Mol. Cell Biol. 20, 3996-4005.

van der Watt, P. J., Maske, C. P., Hendricks, D. T., Parker, M. I., Denny, L.,

Govender, D., Birrer, M. J. and Leaner, V. D. (2009). The Karyopherin proteins,Crm1 and Karyopherin beta1, are overexpressed in cervical cancer and are critical forcancer cell survival and proliferation. Int. J. Cancer 124, 1829-1840.

Van Laer, L., Van Camp, G., van Zuijlen, D., Green, E. D., Verstreken, M.,

Schatteman, I., Van de Heyning, P., Balemans, W., Coucke, P., Greinwald, J. H.

et al. (1997). Refined mapping of a gene for autosomal dominant progressivesensorineural hearing loss (DFNA5) to a 2-cM region, and exclusion of a candidate

gene that is expressed in the cochlea. Eur. J. Hum. Genet. 5, 397-405.

Walde, S. and Kehlenbach, R. H. (2010). The Part and the Whole: functions ofnucleoporins in nucleocytoplasmic transport. Trends Cell Biol. 20, 461-469.

Wen, W., Meinkoth, J. L., Tsien, R. Y. and Taylor, S. S. (1995). Identification of asignal for rapid export of proteins from the nucleus. Cell 82, 463-473.

Wente, S. R. and Rout, M. P. (2010). The Nuclear Pore Complex and NuclearTransport. Cold Spring Harb. Perspect. Biol. 2, a000562.

Wolff, B., Sanglier, J. J. and Wang, Y. (1997). Leptomycin B is an inhibitor of nuclearexport: inhibition of nucleo- cytoplasmic translocation of the human immuno-deficiency virus type 1 (HIV-1) Rev protein and Rev-dependent mRNA. Chem. Biol.

4, 139-147.

Yang, W. and Musser, S. M. (2006). Nuclear import time and transport efficiencydepend on importin beta concentration. J. Cell Biol. 174, 951-961.

Journal of Cell Science 125 (1)154

Journ

alof

Cell

Scie

nce